Abstract： Objective:To analyze the risk factors of neovascular glaucoma (NVG) after 25G pars plana vitrectomy (PPV) in proliferative diabetic retinopathy (PDR).Methods:A retrospective study. From January 2017 to December 2018, 340 PDR patients (340 eyes) with vitreous hemorrhage (VH) who were first treated with PPV in Tianjin Medical University Eye Hospital were included in the study. Among them, 185 were male and 155 were female, with an average age of 55.79±10.82 years. The duration of diabetes was 13.01±7.70 years, the fasting blood glucose was 7.55±2.15 mmol/L. Nineteen patients combined coronary heart disease, and 20 patients combined cerebral infarction. All patients underwent best-corrected visual acuity (BCVA), intraocular pressure (IOP), non-contact fundus examination, and fundus color photographs. BCVA was measured using an international standard Snellen visual acuity chart, and the values were converted to logarithm of the minimum angle of resolution (logMAR) scores for data analysis. The baseline logMAR BCVA was 2.04±0.73, The baseline IOP was 15.45±2.93 mmHg (1 mmHg=0.133 kPa). The duration of VH was 2.98±1.46 months, ranged from 3 weeks to 6 months. Three hundred and forty eyes included 93 eyes of PDR Ⅳ stage (27.35%), 107 eyes of Ⅴ stage (31.47%), and 116 eyes of Ⅵ stage (34.12%), combined tractional retinal detachment (TRD) 83 eyes. All patients underwent 25G standard three channel vitrectomy through the pars plana of the ciliary body. Preoperative anti-VEGF injection was performed in 57 eyes, internal limiting membrane (ILM) peeling in 234 eyes, combined phacoemulsification cataract surgery in 262 eyes and 141 eyes intravitreal anti-VEGF injection at the end of surgery. The patients were followed up for at least 12 months, with an average follow-up time of 10.80±5.79 months. NVG was defined as the presence of neovascularization in the anterior chamber angle or iris with an IOP higher than 21 mmHg after vitrectomy. Kaplan-Meier method and Cox univariate and multivariate regression were used to analyze the relationship between baseline factors, ocular factors, surgical factors and the occurrence of NVG after surgery.Results:Among 340 eyes, 66 eyes (19.41%) developed NVG after vitrectomy during 12 months of observation, NVG occurred from 6 to 335 days after surgery, and the mean period between vitrectomy and developing NVG was 98.00±5.79 days. The incidence of NVG was 11.50%, 15.29% and 20.75%, respectively in the 3rd, 6th and 12th month after PPV. The result of univariate analysis with the Cox regression analysis showed that the development of NVG at 12 months after surgery and age, combined coronary heart disease or cerebral infarction, combined with cataract phacoemulsification, ILM peeling, preoperative anti-VEGF injection had effect on postoperative NVG ( P<0.05). Ocular factors such as PDR staging, combined TRD, preoperative logMAR BCVA, preoperative intraocular pressure, etc. had no effect on the occurrence of NVG after surgery ( P>0.05). Combined cataract phacoemulsification surgery, internal limiting membrane peeling, surgical factors such as intracavity injection of anti-VEGF drugs 3 days before surgery, had an impact on the occurrence of NVG after surgery ( P<0.05). The meaningful variables of the Cox univariate analysis were incorporated into the multivariate Cox proportional hazard model for analysis, and the influencing factors of NVG after surgery were gradually regressed. The results showed that age, coronary heart disease or cerebral infarction, combined with phacoemulsification of cataract, and internal limiting membrane removal during surgery were independent risk predictors of NVG after surgery ( P<0.05). Conclusions:Younger, coronary heart disease or cerebral infarction, combined with cataract phacoemulsification are the risk factors of NVG in PDR patients after PPV. The removal of internal limiting membrane can reduce the incidence of NVG.